We explore bio-inspired training solutions based on the Hebbian principle for deep learning applications in the context of image segmentation tasks.

Setup Environment

This tutorial shows how to setup a python environment with the exact library versions. The tools shown here are asdf and virtualenv

Install asdf

Install asdf using git

git clone https://github.com/asdf-vm/asdf.git ~/.asdf --branch v0.12.0`

Then add to your .bashrc or .zshrc file

. "$HOME/.asdf/asdf.sh"
. "$HOME/.asdf/completions/asdf.bash"

For the changes to have an effect restart the shell.

Install Python

Then install python 3.11.4

asdf plugin add python
asdf install python 3.11.4

Add to .tool-versions

echo "python 3.11.4" >> .tool-versions

Create virtualenv

Upgrade pip and install virtualenv

pip install pip --upgrade
pip install virtualenv

Create new virtual environment

virtualenv venv

Activate the environment

source venv/bin/activate

Install libraries

pip install -r requirements.txt

Usage

Launch experiment with:

python exp.py --config <config> --device <device> --restart

Where:

• <config> is the name of a configuration dictionary, with dotted notation, defined anywhere in your code. For example configs.base.config_base.
• <device> can be cpu, cuda:0, or any device you wish to use for the experiment.
• The flag --restart is optional. If you remove it, you can resume a previously suspended experiment from a checkpoint, if available.

Datasets

Brain MRI images

Requirements

Use requirements.txt

Contacts

Gabriele Lagani: gabriele.lagani@phd.unipi.it